Skip to content

Commit 489c0fd

Browse files
peterschmidt85peterschmidt85
committed
Deploying to gh-pages from @ dstackai/dstack@0424c56 🚀
1 parent ad3b16a commit 489c0fd

26 files changed

Lines changed: 5375 additions & 574 deletions

File tree

‎assets/images/social/blog/smg.png‎

46.3 KB
Loading

‎blog/changelog/index.html‎

Lines changed: 65 additions & 67 deletions
Original file line numberDiff line numberDiff line change
@@ -3923,6 +3923,17 @@
39233923
</label>
39243924
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
39253925

3926+
<li class="md-nav__item">
3927+
<a href="#deploying-sglang-with-pd-disaggregation-via-shepherd-model-gateway" class="md-nav__link">
3928+
<span class="md-ellipsis">
3929+
3930+
Deploying SGLang with PD disaggregation via Shepherd Model Gateway
3931+
3932+
</span>
3933+
</a>
3934+
3935+
</li>
3936+
39263937
<li class="md-nav__item">
39273938
<a href="#infrastructure-orchestration-is-an-agent-skill" class="md-nav__link">
39283939
<span class="md-ellipsis">
@@ -4020,17 +4031,6 @@
40204031
</span>
40214032
</a>
40224033

4023-
</li>
4024-
4025-
<li class="md-nav__item">
4026-
<a href="#supporting-hot-aisle-amd-ai-developer-cloud" class="md-nav__link">
4027-
<span class="md-ellipsis">
4028-
4029-
Supporting Hot Aisle AMD AI Developer Cloud
4030-
4031-
</span>
4032-
</a>
4033-
40344034
</li>
40354035

40364036
</ul>
@@ -4236,6 +4236,17 @@
42364236
</label>
42374237
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
42384238

4239+
<li class="md-nav__item">
4240+
<a href="#deploying-sglang-with-pd-disaggregation-via-shepherd-model-gateway" class="md-nav__link">
4241+
<span class="md-ellipsis">
4242+
4243+
Deploying SGLang with PD disaggregation via Shepherd Model Gateway
4244+
4245+
</span>
4246+
</a>
4247+
4248+
</li>
4249+
42394250
<li class="md-nav__item">
42404251
<a href="#infrastructure-orchestration-is-an-agent-skill" class="md-nav__link">
42414252
<span class="md-ellipsis">
@@ -4333,17 +4344,6 @@
43334344
</span>
43344345
</a>
43354346

4336-
</li>
4337-
4338-
<li class="md-nav__item">
4339-
<a href="#supporting-hot-aisle-amd-ai-developer-cloud" class="md-nav__link">
4340-
<span class="md-ellipsis">
4341-
4342-
Supporting Hot Aisle AMD AI Developer Cloud
4343-
4344-
</span>
4345-
</a>
4346-
43474347
</li>
43484348

43494349
</ul>
@@ -4364,6 +4364,49 @@ <h1 id="changelog">Changelog<a class="headerlink" href="#changelog" title="Perma
43644364
<article class="md-post md-post--excerpt">
43654365
<header class="md-post__header">
43664366

4367+
<div class="md-post__meta md-meta">
4368+
<ul class="md-meta__list">
4369+
<li class="md-meta__item">
4370+
<time datetime="2026-04-29 00:00:00+00:00">April 29, 2026</time></li>
4371+
4372+
<li class="md-meta__item">
4373+
in
4374+
4375+
<a href="./" class="md-meta__link">Changelog</a></li>
4376+
4377+
4378+
4379+
<li class="md-meta__item">
4380+
4381+
3 min read
4382+
4383+
</li>
4384+
4385+
4386+
</ul>
4387+
4388+
</div>
4389+
</header>
4390+
<div class="md-post__content md-typeset">
4391+
<h2 id="deploying-sglang-with-pd-disaggregation-via-shepherd-model-gateway"><a class="toclink" href="../smg/">Deploying SGLang with PD disaggregation via Shepherd Model Gateway</a></h2>
4392+
<p><code>dstack</code> is an open-source control plane that simplifies GPU orchestration for both training and inference — across cloud providers, hardware vendors, and frameworks. Over the past year, we've been steadily making inference a first-class citizen in dstack.</p>
4393+
<p><img src="https://dstack.ai/static-assets/static-assets/images/smg.png" width="630"/></p>
4394+
4395+
4396+
<nav class="md-post__action">
4397+
<a href="../smg/">
4398+
<span>Continue reading</span>
4399+
<span class="icon"><svg viewBox="0 0 13 10" xmlns="http://www.w3.org/2000/svg"><path d="M12.823 4.164L8.954.182a.592.592 0 0 0-.854 0 .635.635 0 0 0 0 .88l2.836 2.92H.604A.614.614 0 0 0 0 4.604c0 .344.27.622.604.622h10.332L8.1 8.146a.635.635 0 0 0 0 .88.594.594 0 0 0 .854 0l3.869-3.982a.635.635 0 0 0 0-.88z" fill-rule="nonzero" fill="currentColor" class="fill-main"></path></svg></span>
4400+
</a>
4401+
</nav>
4402+
4403+
4404+
</div>
4405+
</article>
4406+
4407+
<article class="md-post md-post--excerpt">
4408+
<header class="md-post__header">
4409+
43674410
<div class="md-post__meta md-meta">
43684411
<ul class="md-meta__list">
43694412
<li class="md-meta__item">
@@ -4760,51 +4803,6 @@ <h2 id="introducing-passive-gpu-health-checks"><a class="toclink" href="../gpu-h
47604803
</div>
47614804
</article>
47624805

4763-
<article class="md-post md-post--excerpt">
4764-
<header class="md-post__header">
4765-
4766-
<div class="md-post__meta md-meta">
4767-
<ul class="md-meta__list">
4768-
<li class="md-meta__item">
4769-
<time datetime="2025-08-11 00:00:00+00:00">August 11, 2025</time></li>
4770-
4771-
<li class="md-meta__item">
4772-
in
4773-
4774-
<a href="./" class="md-meta__link">Changelog</a></li>
4775-
4776-
4777-
4778-
<li class="md-meta__item">
4779-
4780-
3 min read
4781-
4782-
</li>
4783-
4784-
4785-
</ul>
4786-
4787-
</div>
4788-
</header>
4789-
<div class="md-post__content md-typeset">
4790-
<h2 id="supporting-hot-aisle-amd-ai-developer-cloud"><a class="toclink" href="../hotaisle/">Supporting Hot Aisle AMD AI Developer Cloud</a></h2>
4791-
<p>As the ecosystem around AMD GPUs matures, developers are looking for easier ways to experiment with ROCm, benchmark new architectures, and run cost-effective workloads—without manual infrastructure setup. </p>
4792-
<p><code>dstack</code> is an open-source orchestrator designed for AI workloads, providing a lightweight, container-native alternative to Kubernetes and Slurm.</p>
4793-
<p><img src="https://dstack.ai/static-assets/static-assets/images/dstack-hotaisle.png" width="630"/></p>
4794-
<p>Today, we’re excited to announce native integration with <a href="https://www.hotaisle.io/">Hot Aisle</a>, an AMD-only GPU neocloud offering VMs and clusters at highly competitive on-demand pricing. </p>
4795-
4796-
4797-
<nav class="md-post__action">
4798-
<a href="../hotaisle/">
4799-
<span>Continue reading</span>
4800-
<span class="icon"><svg viewBox="0 0 13 10" xmlns="http://www.w3.org/2000/svg"><path d="M12.823 4.164L8.954.182a.592.592 0 0 0-.854 0 .635.635 0 0 0 0 .88l2.836 2.92H.604A.614.614 0 0 0 0 4.604c0 .344.27.622.604.622h10.332L8.1 8.146a.635.635 0 0 0 0 .88.594.594 0 0 0 .854 0l3.869-3.982a.635.635 0 0 0 0-.88z" fill-rule="nonzero" fill="currentColor" class="fill-main"></path></svg></span>
4801-
</a>
4802-
</nav>
4803-
4804-
4805-
</div>
4806-
</article>
4807-
48084806

48094807

48104808

‎blog/changelog/page/2/index.html‎

Lines changed: 67 additions & 72 deletions
Original file line numberDiff line numberDiff line change
@@ -3921,6 +3921,17 @@
39213921
</label>
39223922
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
39233923

3924+
<li class="md-nav__item">
3925+
<a href="#supporting-hot-aisle-amd-ai-developer-cloud" class="md-nav__link">
3926+
<span class="md-ellipsis">
3927+
3928+
Supporting Hot Aisle AMD AI Developer Cloud
3929+
3930+
</span>
3931+
</a>
3932+
3933+
</li>
3934+
39243935
<li class="md-nav__item">
39253936
<a href="#rolling-deployment-secrets-files-tenstorrent-and-more" class="md-nav__link">
39263937
<span class="md-ellipsis">
@@ -4018,17 +4029,6 @@
40184029
</span>
40194030
</a>
40204031

4021-
</li>
4022-
4023-
<li class="md-nav__item">
4024-
<a href="#introducing-gpu-blocks-and-proxy-jump-for-ssh-fleets" class="md-nav__link">
4025-
<span class="md-ellipsis">
4026-
4027-
Introducing GPU blocks and proxy jump for SSH fleets
4028-
4029-
</span>
4030-
</a>
4031-
40324032
</li>
40334033

40344034
</ul>
@@ -4234,6 +4234,17 @@
42344234
</label>
42354235
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
42364236

4237+
<li class="md-nav__item">
4238+
<a href="#supporting-hot-aisle-amd-ai-developer-cloud" class="md-nav__link">
4239+
<span class="md-ellipsis">
4240+
4241+
Supporting Hot Aisle AMD AI Developer Cloud
4242+
4243+
</span>
4244+
</a>
4245+
4246+
</li>
4247+
42374248
<li class="md-nav__item">
42384249
<a href="#rolling-deployment-secrets-files-tenstorrent-and-more" class="md-nav__link">
42394250
<span class="md-ellipsis">
@@ -4331,17 +4342,6 @@
43314342
</span>
43324343
</a>
43334344

4334-
</li>
4335-
4336-
<li class="md-nav__item">
4337-
<a href="#introducing-gpu-blocks-and-proxy-jump-for-ssh-fleets" class="md-nav__link">
4338-
<span class="md-ellipsis">
4339-
4340-
Introducing GPU blocks and proxy jump for SSH fleets
4341-
4342-
</span>
4343-
</a>
4344-
43454345
</li>
43464346

43474347
</ul>
@@ -4362,6 +4362,51 @@ <h1 id="changelog">Changelog<a class="headerlink" href="#changelog" title="Perma
43624362
<article class="md-post md-post--excerpt">
43634363
<header class="md-post__header">
43644364

4365+
<div class="md-post__meta md-meta">
4366+
<ul class="md-meta__list">
4367+
<li class="md-meta__item">
4368+
<time datetime="2025-08-11 00:00:00+00:00">August 11, 2025</time></li>
4369+
4370+
<li class="md-meta__item">
4371+
in
4372+
4373+
<a href="../../" class="md-meta__link">Changelog</a></li>
4374+
4375+
4376+
4377+
<li class="md-meta__item">
4378+
4379+
3 min read
4380+
4381+
</li>
4382+
4383+
4384+
</ul>
4385+
4386+
</div>
4387+
</header>
4388+
<div class="md-post__content md-typeset">
4389+
<h2 id="supporting-hot-aisle-amd-ai-developer-cloud"><a class="toclink" href="../../../hotaisle/">Supporting Hot Aisle AMD AI Developer Cloud</a></h2>
4390+
<p>As the ecosystem around AMD GPUs matures, developers are looking for easier ways to experiment with ROCm, benchmark new architectures, and run cost-effective workloads—without manual infrastructure setup. </p>
4391+
<p><code>dstack</code> is an open-source orchestrator designed for AI workloads, providing a lightweight, container-native alternative to Kubernetes and Slurm.</p>
4392+
<p><img src="https://dstack.ai/static-assets/static-assets/images/dstack-hotaisle.png" width="630"/></p>
4393+
<p>Today, we’re excited to announce native integration with <a href="https://www.hotaisle.io/">Hot Aisle</a>, an AMD-only GPU neocloud offering VMs and clusters at highly competitive on-demand pricing. </p>
4394+
4395+
4396+
<nav class="md-post__action">
4397+
<a href="../../../hotaisle/">
4398+
<span>Continue reading</span>
4399+
<span class="icon"><svg viewBox="0 0 13 10" xmlns="http://www.w3.org/2000/svg"><path d="M12.823 4.164L8.954.182a.592.592 0 0 0-.854 0 .635.635 0 0 0 0 .88l2.836 2.92H.604A.614.614 0 0 0 0 4.604c0 .344.27.622.604.622h10.332L8.1 8.146a.635.635 0 0 0 0 .88.594.594 0 0 0 .854 0l3.869-3.982a.635.635 0 0 0 0-.88z" fill-rule="nonzero" fill="currentColor" class="fill-main"></path></svg></span>
4400+
</a>
4401+
</nav>
4402+
4403+
4404+
</div>
4405+
</article>
4406+
4407+
<article class="md-post md-post--excerpt">
4408+
<header class="md-post__header">
4409+
43654410
<div class="md-post__meta md-meta">
43664411
<ul class="md-meta__list">
43674412
<li class="md-meta__item">
@@ -4808,56 +4853,6 @@ <h2 id="auto-shutdown-for-inactive-dev-environmentsno-idle-gpus"><a class="tocli
48084853
</div>
48094854
</article>
48104855

4811-
<article class="md-post md-post--excerpt">
4812-
<header class="md-post__header">
4813-
4814-
<div class="md-post__meta md-meta">
4815-
<ul class="md-meta__list">
4816-
<li class="md-meta__item">
4817-
<time datetime="2025-02-18 00:00:00+00:00">February 18, 2025</time></li>
4818-
4819-
<li class="md-meta__item">
4820-
in
4821-
4822-
<a href="../../" class="md-meta__link">Changelog</a></li>
4823-
4824-
4825-
4826-
<li class="md-meta__item">
4827-
4828-
4 min read
4829-
4830-
</li>
4831-
4832-
4833-
</ul>
4834-
4835-
</div>
4836-
</header>
4837-
<div class="md-post__content md-typeset">
4838-
<h2 id="introducing-gpu-blocks-and-proxy-jump-for-ssh-fleets"><a class="toclink" href="../../../gpu-blocks-and-proxy-jump/">Introducing GPU blocks and proxy jump for SSH fleets</a></h2>
4839-
<p>Recent breakthroughs in open-source AI have made AI infrastructure accessible beyond public clouds, driving demand for
4840-
running AI workloads in on-premises data centers and private clouds.
4841-
This shift offers organizations both high-performant clusters and flexibility and control.</p>
4842-
<p>However, Kubernetes, while a popular choice for traditional deployments, is often too complex and low-level to address
4843-
the needs of AI teams.</p>
4844-
<p>Originally, <code>dstack</code> was focused on public clouds. With the new release, <code>dstack</code>
4845-
extends support to data centers and private clouds, offering a simpler, AI-native solution that replaces Kubernetes and
4846-
Slurm.</p>
4847-
<p><img src="https://dstack.ai/static-assets/static-assets/images/data-centers-and-private-clouds.png" width="630"/></p>
4848-
4849-
4850-
<nav class="md-post__action">
4851-
<a href="../../../gpu-blocks-and-proxy-jump/">
4852-
<span>Continue reading</span>
4853-
<span class="icon"><svg viewBox="0 0 13 10" xmlns="http://www.w3.org/2000/svg"><path d="M12.823 4.164L8.954.182a.592.592 0 0 0-.854 0 .635.635 0 0 0 0 .88l2.836 2.92H.604A.614.614 0 0 0 0 4.604c0 .344.27.622.604.622h10.332L8.1 8.146a.635.635 0 0 0 0 .88.594.594 0 0 0 .854 0l3.869-3.982a.635.635 0 0 0 0-.88z" fill-rule="nonzero" fill="currentColor" class="fill-main"></path></svg></span>
4854-
</a>
4855-
</nav>
4856-
4857-
4858-
</div>
4859-
</article>
4860-
48614856

48624857

48634858

0 commit comments

Comments
 (0)